Regional agricultural planting optimization method based on uncertainty

Abstract

The invention discloses a regional agricultural planting optimization method based on uncertainty, belongs to the technical field of agricultural decision optimization, and constructs a regional agricultural planting optimization model considering uncertainty, and the model takes planting density as a decision variable and takes economic benefit maximization and resource conflict entropy maximization as targets. Considering that the yield meets the minimum demand and the planting density is non-negative constraint; a double-domain elastic triangular set method is provided to fuzzify the uncertainty parameters, and the parameters are determined through a double-domain tradeoff center method; a target upper limit elastic expansion method and a target lower limit elastic reduction method are adopted to realize upper and lower limit bidirectional dynamic adjustment of a regional agricultural planting optimization model target, and a regional agricultural optimization planting scheme is solved through an adaptive multi-step gradient optimization algorithm. According to the method, the problems of parameter solidification and insufficient decision-making elasticity of a current traditional planting planning model are solved, and the method is of great significance to regional agricultural planting density optimization.

Description

Technical Field

[0001] The present invention relates to the technical field of agricultural planting optimization, and in particular to a regional agricultural planting optimization method. Background Art

[0002] Traditional agricultural planting optimization methods often focus on allocating land resources and irrigation strategies, while neglecting the control of planting density. Existing methods have significant shortcomings in optimizing planting density: First, a lack of basis for density setting often leads to problems in actual planting, such as excessively high density leading to reduced yields due to nutrient competition, or excessively low density leading to reduced land utilization. Second, these static control models struggle to adapt to the dynamic demands of the crop growth cycle and cannot flexibly adjust to real-time factors such as soil moisture and climatic conditions, ultimately leading to inefficient resource utilization and diminished economic returns. Traditional planting density optimization also has limitations: The goal setting is too simplistic, focusing on single-dimensional optimization such as maximizing yield or minimizing cost. The control mechanism lacks flexibility, failing to establish a two-way dynamic adjustment system with upper and lower limits. Furthermore, the method lacks adaptability, making it difficult to cope with complex and volatile external environments. When faced with uncertainties such as volatile market prices or frequent extreme weather events, traditional static optimization models exhibit the drawback of being unable to fully unleash production potential by exceeding the upper density limit during favorable market conditions, nor can they effectively mitigate losses by tightening the lower density limit during risk warnings. In summary, there is an urgent need for a new optimization method that integrates the upper and lower limit elasticity strategies with planting density as the target, so as to achieve refined and adaptive management of agricultural production by dynamically balancing planting density. Summary of the Invention

[0003] In response to the shortcomings of the existing technology, the present invention provides a regional agricultural planting optimization method based on uncertainty. Through the dual-domain elastic triangle set method and the dual-domain trade-off center method, the problem of parameter uncertainty in traditional agricultural planting optimization methods is solved; through the target upper limit elastic expansion method and the target lower limit elastic reduction method, the decision maker's preference for results is solved, and at the same time, the maximization of economic benefits and the maximization of resource conflict entropy can be achieved.

[0004] The object of the present invention is achieved as follows: a regional agricultural planting optimization method based on uncertainty, comprising the following steps:

[0005] S1: Determine the climate parameters, crop economic parameters and agronomic parameters within the input region;

[0006] S2: Construct a regional agricultural planting optimization model considering uncertainty; take the planting density x of the jth crop in the i-th water resource area as ijAs decision variables, the maximum economic benefit maxF1 and the maximum resource conflict entropy maxF2 are taken as objective functions, and the yield meets the minimum demand and the planting density is non-negative as constraints. Considering the uncertainty of rainfall, accumulated temperature, crop economics and yield parameters in the region, a regional agricultural planting optimization model is constructed.

[0007] S3: Parameter fuzzification and defuzzification: The uncertain parameters in S2 are fuzzified using a dual-domain elastic triangular set method that includes support intensity function and opposition intensity function, and the uncertain parameters are defuzzified using a dual-domain trade-off center method.

[0008] S4: Model solution; Taking the maximum satisfaction value max λ as the goal, the upper limit elastic expansion method and the lower limit elastic reduction method are used to realize the two-way dynamic adjustment of the upper and lower limits of the regional agricultural planting optimization model, forming a transformation model based on the upper limit elastic expansion method and a transformation model based on the lower limit elastic reduction method respectively; the adaptive multi-step gradient optimization algorithm is used to solve, and the planting density x of the jth crop in the i-th water resource area of the above two transformation models is obtained respectively. ij ;

[0009] S5: Output regional agricultural optimized planting plan.

[0010] Furthermore, the climate parameters, crop economic parameters and agronomic parameters in the region in S1 are as follows: the climate parameters include: the rainfall of each crop in each season, the average daily wind speed U at a height of 2 meters in the kth growth period in the region, ik and temperature T ik , the slope of the saturated water vapor pressure curve Δ in the kth growth period in the region ik , the saturated water vapor pressure e in the kth growth period in the region sik and the actual water vapor pressure e aik ; Crop economic parameters include: market price of each crop in the region P ij , the production cost of each crop in the region C ij ; Agronomic parameters include: the yield of each crop in the region Y ij , accumulated temperature of each crop in each quarter, crop coefficient K of each crop j , the net radiation RO of the crop surface in the kth growth period in the region ik , the soil heat flux density G of the kth growth period in the region ik , the psychrometric constant γ of the kth growth period in the region ik ;

[0011] Among them, the yield of each crop in the region is Y ij Will increase with the increase of planting density, and use the constraint line method to fit the crop yield Y ijThe relationship between crop density and crop yield is fitted as follows: First, collect more than 500 sets of field data and draw a scatter plot with the x-axis representing crop density and the y-axis representing crop yield. Then, identify the data points in the scatter plot with the highest yield under different crop planting densities. Then, use the exponential regression method to fit the boundary line, and obtain the following equation:

[0012] ;

[0013] Where a ij 、b ij and c ij is the yield fitting coefficient of the jth crop in the i-th water resource area.

[0014] Furthermore, the objective function of S2 is as follows:

[0015] Goal 1: Economic benefit goal:

[0016] ;

[0017] Goal 2: Resource Conflict Entropy Goal:

[0018] ;

[0019] Where i is the water resource area, i=1~m, m is the total number of water resource areas; j is the crop type, j=1~n, n is the total number of crop types; P ij and C ij are the price and production cost of the jth crop in the i-th water resource area; Y ij is the yield of the jth crop in the i-th water resource area; ξ is the weight coefficient; Q i is the total amount of available water resources in the i-th water resource area; and is the price and production cost of the jth crop in the i-th water resource area after fuzzification; is the yield of the jth crop in the i-th water resource area after fuzzy processing; is the seasonal rainfall variation coefficient of the jth crop after fuzzy processing; is the quarterly accumulated temperature deviation of the jth crop after fuzzy processing; R j is the seasonal rainfall variation coefficient of the jth crop, D j is the cumulative deviation of quarterly accumulated temperature of the jth crop, W ij is the water requirement per unit area of the jth crop in the i-th water resource area;

[0020] The output of S2 satisfies the minimum demand constraint as follows:

[0021] ;

[0022] In the formula, EY ij represents the yield per plant of the jth crop in the i-th water resource area; Q j represents the minimum demand for the jth crop;

[0023] The non-negative constraint condition of the planting density of S2 is as follows:

[0024] .

[0025] Furthermore, the uncertainty parameters in S3 are regional rainfall, accumulated temperature, crop economic and yield parameters, including: seasonal rainfall variation coefficient R of each crop j , the cumulative deviation of quarterly accumulated temperature of each crop D j , the market price of each crop in the region P ij , the production cost of each crop in the region C ij , the yield of each crop in the region Y ij .

[0026] Furthermore, the dual-domain elastic triangle set method in S3 is used to calculate the support strength and opposition strength of uncertainty parameters. The specific calculation formula is as follows:

[0027] ;

[0028] ;

[0029] Where, is a dual-domain elastic fuzzy set The support strength of the uncertainty parameter x; is a dual-domain elastic fuzzy set The opposition strength of the uncertainty parameter x; represents the dual-domain elastic fuzzy set of the above uncertainty parameters; x represents the uncertainty parameter; μ and θ represent the support coefficient and opposition coefficient respectively; a1 is the minimum value of the uncertainty parameter; a2 is the average value of the uncertainty parameter; a3 is the maximum value of the uncertainty parameter; a1 ’ is the lower quartile of the uncertainty parameter; a3 ’ The upper quartile of the uncertainty parameter.

[0030] Furthermore, the dual-domain weighted center method in S3 is achieved by 、 Will Convert to a fixed value , and its calculation formula is:

[0031] ;.

[0032] Furthermore, the target upper limit elastic expansion method in S4 uses a nonlinear dynamic adjustment method to expand the objective function of S2 in a favorable direction and convert it into a constraint condition, as follows:

[0033] The upper limit elastic expansion method converts the maximum value target into a constraint condition F op v (x ij ):

[0034] ;

[0035] The target lower limit elastic reduction method in S4 uses a nonlinear dynamic adjustment method to dynamically expand the objective function of S2 in an unfavorable direction and convert it into a constraint condition, as follows:

[0036] The lower bound elastic reduction method transforms the maximum target into the constraint condition F pp v (x ij ):

[0037] ;

[0038] Where, F v,max (x ij ) and F v,min (x ij ) are the vth target F v (x ij ), v=1-l, l is the number of targets; u is the exponent for adjusting nonlinearity, ; η is the degree of inclination, ; Δ v For the vth target F v (x ij )'s maximum and minimum values.

[0039] Furthermore, the conversion model based on the upper limit elastic expansion method in S4 is specifically as follows:

[0040] Objective function: Max λ;

[0041] Constraints: ;

[0042] ;

[0043] ;

[0044] .

[0045] Furthermore, the conversion model based on the lower bound elasticity reduction method in S4 is specifically as follows:

[0046] Objective function: Max λ;

[0047] Constraints: ;

[0048] ;

[0049] ;

[0050] .

[0051] Furthermore, the adaptive multi-step gradient optimization algorithm in S4 is used to solve the conversion model based on the upper limit elastic expansion method and the conversion model based on the lower limit elastic reduction method. The specific steps are as follows:

[0052] Step 1: Model initialization and step size rule;

[0053] Step 2: Calculate the gradient;

[0054] Step 3: Activate constraint judgment and construct the reduced gradient;

[0055] Step 4: Direction search and step size selection;

[0056] Step 5: Feasibility revision;

[0057] Step 6: Convergent judgment.

[0058] Compared with the prior art, the present invention has the following beneficial effects:

[0059] 1. This invention uses planting density as the core decision variable to accurately quantify crop allocation, breaking through the limitations of traditional models that rely on fixed planting scales;

[0060] 2. This invention adopts the dual-domain elastic triangulation method and the dual-domain trade-off center method, which fully considers the parameter uncertainty caused by climate change and market fluctuations in actual agricultural production, and overcomes the problem that traditional agricultural planting optimization determination models cannot effectively deal with the uncertainty factors in actual production;

[0061] 3. The present invention also adopts the target upper limit elastic expansion method and the target lower limit elastic reduction method to realize the two-way dynamic adjustment of the upper and lower limits of the regional agricultural planting optimization model target, overcoming the situation that the traditional static optimization model is often unable to perform effective dynamic adjustment under extreme circumstances, resulting in the optimization plan being out of touch with actual decision-making needs.

[0062] 4. The present invention is of great significance to improving agricultural production efficiency and quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0064] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0065] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0066] like Figure 1 The uncertainty-based regional agricultural planting optimization method shown includes the following steps:

[0067] S1: Determine the climate parameters, crop economic parameters, and agronomic parameters within the input region. Specifically, the climate parameters include: rainfall for each crop in each season, the average daily wind speed U at a height of 2 meters during the kth growth period within the region. ik and temperature T ik , the slope of the saturated water vapor pressure curve Δ in the kth growth period in the region ik , the saturated water vapor pressure e in the kth growth period in the region sik and the actual water vapor pressure e aik ; Crop economic parameters include: market price of each crop in the region P ij , the production cost of each crop in the region C ij ; Agronomic parameters include: the yield of each crop in the region Y ij , accumulated temperature of each crop in each quarter, crop coefficient K of each crop j , the net radiation RO of the crop surface in the kth growth period in the region ik , the soil heat flux density G of the kth growth period in the region ik , the psychrometric constant γ of the kth growth period in the region ik The yield fitting coefficients of the four crops selected in the actual example are shown in Table 1. The yield fitting coefficients of the four crops in the 21 water resource areas are the same.

[0068] Table 1: Yield fitting coefficients of four crops

[0069]

[0070] S2: Construct a regional agricultural planting optimization model considering uncertainty; take the planting density x of the jth crop in the i-th water resource area as ij As decision variables, the maximum economic benefit maxF1 and the maximum resource conflict entropy maxF2 are taken as objective functions, and the yield meets the minimum demand and the planting density is non-negative as constraints. Considering the uncertainty of rainfall, accumulated temperature, crop economics and yield parameters in the region, a regional agricultural planting optimization model is constructed.

[0071] S2-1: Uncertainty parameters: regional rainfall, accumulated temperature, crop economic and yield parameters, including: seasonal rainfall coefficient of variation R of each crop j , the cumulative deviation of quarterly accumulated temperature of each crop D j , the market price of each crop in the region P ij , the production cost of each crop in the region C ij , the yield of each crop in the region Y ij .

[0072] S2-2: Construct a regional agricultural planting optimization model. The specific model is as follows:

[0073] Goal 1: Economic benefit goal

[0074] (1)

[0075] Goal 2: Resource Conflict Entropy Goal (2)

[0076] Where i is the water resource area, i=1~m, m is the total number of water resource areas; j is the crop type, j=1~n, n is the total number of crop types; P ij (RMB·plant / ha) and C ij (RMB·plant / ha) are the price and production cost of the jth crop in the i-th water resource area; Y ij (kg·plant / ha) is the yield of the jth crop in the i-th water resource area; ξ is the weight coefficient; Q i (m 3 / ha) is the total available water resources in the i-th water resource area; (RMB·strain / ha) and (RMB·plant / ha) is the price and production cost of the jth crop in the i-th water resource area after fuzzification; (kg·plant / ha) is the yield of the jth crop in the i-th water resource area after fuzzification; is the seasonal rainfall variation coefficient of the jth crop after fuzzy processing; is the quarterly accumulated temperature deviation of the jth crop after fuzzy processing; Rj is the seasonal rainfall variation coefficient of the jth crop, and the formula is:

[0077] (3)

[0078] Where R σ j is the standard deviation of seasonal rainfall for the jth crop, R ave j is the mean seasonal rainfall of the jth crop; D j is the cumulative deviation of quarterly accumulated temperature of the jth crop, and the formula is:

[0079] (4)

[0080] Where D a j is the actual accumulated temperature of the jth crop, D ave j is the average accumulated temperature of the jth crop; W ij (m 3 / ha) is the water requirement per unit area of the jth crop in the i-th water resource zone, and the formula is:

[0081] (5)

[0082] (6)

[0083] Where k is the crop growth cycle, k=1~h, h is the number of crop growth cycles; ET ijk (mm / d) represents the evapotranspiration of the jth crop in the i-th water resource area during the k-th growth period; K j represents the crop coefficient of the jth crop; S ij represents the crop area of the jth crop in the i-th water resource area; Δ ik (kPa / °C) represents the slope of the saturated water vapor pressure curve of the i-th water resource area in the k-th growth period; RO ik (MJ / (m 2 ·day)) represents the net radiation of the crop surface in the i-th water resource area during the k-th growth period; G ik (MJ / (m 2 d)) represents the soil heat flux density of the i-th water resource area in the k-th growth period; γ ik (kPa / °C) represents the psychrometric constant of the i-th water resource area in the k-th growth period; U ik (m / s) represents the daily average wind speed at a height of 2 meters in the i-th water resource area during the k-th growth period; T ik (°C) represents the air temperature at a height of 2 meters in the i-th water resource area during the k-th growth period; esik (kPa) represents the saturated water vapor pressure of the i-th water resource area in the k-th growth period; e aik (kPa) represents the actual water vapor pressure of the i-th water resource area during the k-th growth period.

[0084] Constraint 1: Output meets minimum demand:

[0085] (7)

[0086] Constraint 2: Planting density is non-negative:

[0087] (8)

[0088] In the formula, EY ij (kg / plant) represents the yield per plant of the jth crop in the i-th water resource area; Q j (kg) represents the minimum demand for the jth crop.

[0089] S3: Parameter fuzzification and defuzzification. The uncertain parameters in S2 are fuzzified using a dual-domain elastic triangular set method that includes support intensity function and opposition intensity function, and the uncertain parameters are defuzzified using a dual-domain trade-off center method.

[0090] S3-1: The support and opposition strengths of uncertainty parameters are calculated using the dual-domain elastic triangle set method. The specific calculation formula is as follows:

[0091] (9)

[0092] (10)

[0093] Where, is a dual-domain elastic fuzzy set The support strength of the uncertainty parameter x; is a dual-domain elastic fuzzy set The opposition strength of the uncertainty parameter x; represents the dual-domain elastic fuzzy set of the above uncertainty parameters; x represents the uncertainty parameter; μ and θ represent the support coefficient and opposition coefficient respectively; a1 is the minimum value of the uncertainty parameter; a2 is the average value of the uncertainty parameter; a3 is the maximum value of the uncertainty parameter; a1 ’ is the lower quartile of the uncertainty parameter; a3 ’ The upper quartile of the uncertainty parameter.

[0094] S3-2: Using the dual-domain trade-off center method Convert to a fixed value , the specific calculation formula is:

[0095] (11)

[0096] S4: Model solution; Taking the maximum satisfaction value max λ as the goal, the upper limit elastic expansion method and the lower limit elastic reduction method are used to realize the two-way dynamic adjustment of the upper and lower limits of the regional agricultural planting optimization model, forming a transformation model based on the upper limit elastic expansion method and a transformation model based on the lower limit elastic reduction method respectively; the adaptive multi-step gradient optimization algorithm is used to solve, and the planting density x of the jth crop in the i-th water resource area of the above two transformation models is obtained respectively. ij .

[0097] S4-1: This invention uses a nonlinear dynamic adjustment method to expand the objective function of S2 toward the upper and lower limits, respectively, in order to significantly improve the optimization accuracy of the model. Traditional methods use linear adjustment methods, but in actual applications, there are cases where decision makers have preferences for certain objectives, and linear adjustment methods cannot solve this situation. Therefore, we use a nonlinear dynamic adjustment method, the specific formula is as follows:

[0098] The upper limit elastic expansion method converts the maximum value target into a constraint condition F op v (x ij ):

[0099] (12)

[0100] The lower bound elastic reduction method transforms the maximum target into the constraint condition F pp v (x ij ):

[0101] (13)

[0102] Where, F v，max (x ij ) and F v,min (x ij ) are the vth target F v (x ij ), v=1-l, l is the number of targets; u is the exponent for adjusting nonlinearity, ; η is the degree of inclination, ; Δ v For the vth target F v (x ij )'s maximum and minimum values.

[0103] S4-2: Convert the multi-objective model into a single-objective model while retaining the original constraints, forming a conversion model based on the upper limit elastic expansion method and a conversion model based on the lower limit elastic reduction method, as follows:

[0104] Conversion model based on upper limit elastic expansion method:

[0105] Objective function: Max λ(14)

[0106] Constraints: (15)

[0107] (16)

[0108] (17)

[0109] (18).

[0110] Conversion model based on lower bound elasticity reduction method:

[0111] Objective function: Max λ(19)

[0112] Constraints: (20)

[0113] (twenty one)

[0114] (twenty two)

[0115] (twenty three).

[0116] S4-3: Adaptive multi-step gradient optimization algorithm is used to solve the transformation model based on the upper limit elastic expansion method and the transformation model based on the lower limit elastic reduction method to obtain the planting density x of the jth crop in the i-th water resource area. ij , the specific steps are as follows:

[0117] Step 1: Model initialization and step size rules:

[0118] Parameter n: Use fine adjustment, increase or decrease in steps of 0.1.

[0119] Parameter u: Adjust by integer units, increase or decrease in steps of 1.

[0120] Parameter Δ v : Calculation method is Δ v = maximum or minimum value of the target × percentage, adjust the percentage in steps of 1% to adjust Δ v The value of .

[0121] Select an initial feasible point , the initial feasible point satisfies the constraints.

[0122] Step 2: Calculate the gradient:

[0123] First calculate the objective function gradient , is a vector of the form:

[0124] (twenty four)

[0125] Where 1 represents the partial derivative of the objective function with respect to variable λ only, and 0 represents the partial derivative with respect to other variables.

[0126] Then calculate the constrained gradient, for 、 、 、 The four nonlinear constraints, the constraint gradients are 、 、 、 ,by For example, the formula is as follows:

[0127] (25)

[0128] Where, , ,… Indicates that the first g components are F od v (x ij ) for x ij The partial derivative of -1 represents the partial derivative of the last component with respect to λ, and g is x ij The number of remaining nonlinear constraints is calculated as above.

[0129] For linear constraints , the constrained gradient formula is:

[0130] (26)

[0131] Where, , ,… Represents the first g components of x ij Find the partial derivative, where 0 represents the partial derivative of the last component with respect to λ.

[0132] Step 3: Activate constraint judgment and construct constrained gradient

[0133] If the constraints satisfy the equation, such as: or or or or , it is called activation constraint. If the above constraint is activation constraint, then the gradient formula is reduced to For example:

[0134] (27)

[0135] Where, is the Lagrange multiplier.

[0136] Step 4: Direction search and step size selection

[0137] Search direction d, if no activation constraint is present, along That is, increase the direction of λ; if there is an activation constraint, follow the negative direction of the gradient. Through line search, find the largest α such that and , , , .

[0138] Step 5: Feasibility revision

[0139] If after iteration x ij e+1 If some constraints are not satisfied, use projection method or backtracking line search to adjust α and then recalculate x ij e+1 Return it to the feasible region.

[0140] Step 6: Convergence Judgment

[0141] When the gradient norm is reduced The change in λ is less than the preset accuracy threshold. The process terminates when the value is less than the preset precision value. If the previous conditions are not met, the process returns to the first step and repeats the above process until the optimal planting density is obtained.

[0142] S5: Output regional agricultural optimized planting plan.

[0143] The present invention is further illustrated below with reference to specific examples: Taking Jiangsu Province as an example, the technical method of the present invention is used to study the optimization of agricultural planting density in the study area.

[0144] According to the national economic development plan and the comprehensive utilization plan of water resources, Jiangsu Province can be divided into 21 fourth-level water resource zones, namely: Z1: Fengpei District; Z2: Luomahu District; Z3: Ganyu District; Z4: Yibei District; Z5: Yinan District; Z6: Anhe District; Z7: Xuyi District; Z8: Qubei District; Z9: Gaobao Lake District; Z10: Lixiahe Central Area; Z11: Doubei District; Z12: Dounan District; Z13: Yiliu District; Z14: Qinhuaihe District; Z15: Gucheng Shijiu District; Z16: Huxi District; Z17: Tongnan Riverside District (Yangzhou); Z18: Tongnan Riverside District (Tongzhou); Z19: Wucheng Xiyu District; Z20: Yangcheng Dianmao District; Z21: Punan District, that is, i=21. Four major rainy season crops were selected: rice (Nanjing 9108), corn (Suyu 29), peanut (Xuhua 13), and cotton (Sikang 1), i.e., j = 4. The rainfall of each crop in each season in Jiangsu Province from 2000 to 2020, the accumulated temperature of each crop in each season, and the slope of the saturated water vapor pressure curve Δ of the kth growth cycle in the region were collected from the China Meteorological Data Network. ik , the net radiation RO of the crop surface in the kth growth period in the region ik , the psychrometric constant γ of the kth growth period in the region ik , the average daily wind speed U at a height of 2 meters in the kth growth period in the region ik and temperature T ik , the saturated water vapor pressure e in the kth growth period in the region sik and the actual water vapor pressure e aik The crop coefficient K of each crop was collected from the FAO database, the International Soil Reference Information Center and the China Soil Science Database. j and the soil heat flux density G of the kth growth period in the region ik The market prices of various crops in the region were collected from the Jiangsu Statistical Yearbook from 2000 to 2020. ij and output Y ij The production cost C of each crop in the region of the above four crops in Jiangsu Province was collected from the "National Agricultural Product Cost Data Compilation" from 2000 to 2020. ij According to the uncertainty-based regional agricultural planting optimization method proposed in this patent, the optimized agricultural planting plan for each water resource area in Jiangsu Province was obtained. The results are shown in Tables 2 and 3.

[0145] Table 2: Agricultural optimization planting plan for each water resource area in Jiangsu Province based on the upper limit elastic expansion method according to the embodiment of the present invention

[0146]

[0147]

[0148] Table 3: Agricultural optimization planting plan for each water resource area in Jiangsu Province based on the lower bound elastic expansion method according to the embodiment of the present invention

[0149]

[0150]

[0151]

[0152] The above embodiments are only intended to help understand the method and core concept of the present invention. It should be noted that, without departing from the principles of the present invention, a number of improvements and modifications may be made to the present invention by those skilled in the art, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A regional agricultural planting optimization method based on uncertainty, characterized in that: The following steps are involved: S1: Determine the climate parameters, crop economic parameters and agronomic parameters within the input region; S2: Construct a regional agricultural planting optimization model considering uncertainty; take the planting density x of the jth crop in the i-th water resource area as ij As decision variables, the maximum economic benefit maxF1 and the maximum resource conflict entropy maxF2 are taken as objective functions, and the yield meets the minimum demand and the planting density is non-negative as constraints. Considering the uncertainty of rainfall, accumulated temperature, crop economics and yield parameters in the region, a regional agricultural planting optimization model is constructed. S3: Parameter fuzzification and defuzzification: The uncertain parameters in S2 are fuzzified using a dual-domain elastic triangular set method that includes support intensity function and opposition intensity function, and the uncertain parameters are defuzzified using a dual-domain trade-off center method. S4: Model solution; Taking the maximum satisfaction value max λ as the goal, the upper limit elastic expansion method and the lower limit elastic reduction method are used to realize the two-way dynamic adjustment of the upper and lower limits of the regional agricultural planting optimization model, forming a transformation model based on the upper limit elastic expansion method and a transformation model based on the lower limit elastic reduction method respectively; the adaptive multi-step gradient optimization algorithm is used to solve, and the planting density x of the jth crop in the i-th water resource area of the above two transformation models is obtained respectively. ij ; S5: Output regional agricultural optimized planting plan.

2. The regional agricultural planting optimization method based on uncertainty according to claim 1 is characterized in that: The climate parameters, crop economic parameters and agronomic parameters in the region in S1 are as follows: the climate parameters include: the rainfall of each crop in each season, the average daily wind speed U at a height of 2 meters in the kth growth period in the region ik and temperature T ik , the slope of the saturated water vapor pressure curve Δ in the kth growth period in the region ik , the saturated water vapor pressure e in the kth growth period in the region sik and the actual water vapor pressure e aik ; Crop economic parameters include: market price of each crop in the region P ij , the production cost of each crop in the region C ij ; Agronomic parameters include: the yield of each crop in the region Y ij , accumulated temperature of each crop in each quarter, crop coefficient K of each crop j , the net radiation RO of the crop surface in the kth growth period in the region ik , the soil heat flux density G of the kth growth period in the region ik , the psychrometric constant γ of the kth growth period in the region ik ; Among them, the yield of each crop in the region is Y ij Will increase with the increase of planting density, and use the constraint line method to fit the crop yield Y ij The relationship between crop density and crop yield is fitted as follows: First, collect more than 500 sets of field data and draw a scatter plot with the x-axis representing crop density and the y-axis representing crop yield. Then, identify the data points in the scatter plot with the highest yield under different crop planting densities. Then, use the exponential regression method to fit the boundary line, and obtain the following equation: ； Where a ij 、b ij and c ij is the yield fitting coefficient of the jth crop in the i-th water resource area.

3. The regional agricultural planting optimization method based on uncertainty according to claim 2 is characterized in that: The objective function of S2 is as follows: Goal 1: Economic benefit goal: ； Goal 2: Resource Conflict Entropy Goal: ； Where i is the water resource area, i=1~m, m is the total number of water resource areas; j is the crop type, j=1~n, n is the total number of crop types; P ij and C ij are the price and production cost of the jth crop in the i-th water resource area; Y ij is the yield of the jth crop in the i-th water resource area; ξ is the weight coefficient; Q i is the total amount of available water resources in the i-th water resource area; and is the price and production cost of the jth crop in the i-th water resource area after fuzzification; is the yield of the jth crop in the i-th water resource area after fuzzy processing; is the seasonal rainfall variation coefficient of the jth crop after fuzzy processing; is the quarterly accumulated temperature deviation of the jth crop after fuzzy processing; R j is the seasonal rainfall variation coefficient of the jth crop, D j is the cumulative deviation of quarterly accumulated temperature of the jth crop, W ij is the water requirement per unit area of the jth crop in the i-th water resource area; The output of S2 satisfies the minimum demand constraint as follows: ； In the formula, EY ij represents the yield per plant of the jth crop in the i-th water resource area; Q j represents the minimum demand for the jth crop; The non-negative constraint condition of the planting density of S2 is as follows: 。 4. The regional agricultural planting optimization method based on uncertainty according to claim 3 is characterized in that: The uncertainty parameters in S3 are the rainfall, accumulated temperature, crop economic and yield parameters in the region, including: seasonal rainfall variation coefficient R of each crop j , the cumulative deviation of quarterly accumulated temperature of each crop D j , the market price of each crop in the region P ij , the production cost of each crop in the region C ij , the yield of each crop in the region Y ij .

5. The regional agricultural planting optimization method based on uncertainty according to claim 4 is characterized in that: The dual-domain elastic triangle set method in S3 is used to calculate the support strength and opposition strength of uncertainty parameters. The specific calculation formula is as follows: ； ； Where, is a dual-domain elastic fuzzy set The support strength of the uncertainty parameter x; is a dual-domain elastic fuzzy set The opposition strength of the uncertainty parameter x; represents the dual-domain elastic fuzzy set of the above uncertainty parameters; x represents the uncertainty parameter; μ and θ represent the support coefficient and opposition coefficient respectively; a1 is the minimum value of the uncertainty parameter; a2 is the average value of the uncertainty parameter; a3 is the maximum value of the uncertainty parameter; a1 ’ is the lower quartile of the uncertainty parameter; a3 ’ The upper quartile of the uncertainty parameter.

6. The regional agricultural planting optimization method based on uncertainty according to claim 5 is characterized in that: The dual-domain trade-off center method in S3 is achieved by 、 Will Convert to a fixed value , and its calculation formula is: ；。 7. The regional agricultural planting optimization method based on uncertainty according to claim 6 is characterized in that: The target upper limit elastic expansion method in S4 uses a nonlinear dynamic adjustment method to expand the objective function of S2 in a favorable direction and convert it into a constraint condition, as follows: The upper limit elastic expansion method converts the maximum value target into a constraint condition F op v (x ij ): ； The target lower limit elastic reduction method in S4 uses a nonlinear dynamic adjustment method to dynamically expand the objective function of S2 in an unfavorable direction and convert it into a constraint condition, as follows: The lower bound elastic reduction method transforms the maximum value target into the constraint condition F pp v (x ij ): ； Where, F v,max (x ij ) and F v,min (x ij ) are the vth target F v (x ij ), v=1-l, l is the number of targets; u is the exponent for adjusting nonlinearity, ; η is the degree of inclination, ; Δ v For the vth target F v (x ij )'s maximum and minimum values.

8. The regional agricultural planting optimization method based on uncertainty according to claim 7 is characterized in that: The conversion model based on the upper limit elastic expansion method in S4 is specifically as follows: Objective function: Max λ; Constraints: ; ； ； 。 9. The regional agricultural planting optimization method based on uncertainty according to claim 7 is characterized in that: The conversion model based on the lower bound elasticity reduction method in S4 is specifically as follows: Objective function: Max λ; Constraints: ; ； ； 。 10. The regional agricultural planting optimization method based on uncertainty according to claim 7, characterized in that: The adaptive multi-step gradient optimization algorithm in S4 is used to solve the conversion model based on the upper limit elastic expansion method and the conversion model based on the lower limit elastic reduction method. The specific steps are as follows: Step 1: Model initialization and step size rule; Step 2: Calculate the gradient; Step 3: Activate constraint judgment and construct the reduced gradient; Step 4: Direction search and step size selection; Step 5: Feasibility revision; Step 6: Convergent judgment.

Images